Embed Size (px)
Citation preview
LECTURE 12
TSUNAMI W ARNING :
ALGORITHMS and
MODERN IMPR OVEMENTS
© Northwestern University, 2007
TSUNAMI WARNING: IDEA and CHALLENGES
• Tsunamis are much slower (220 m/s) than seismic wav es (3to 10 km/s).
• Hence, it should be possible to provide warning, at least inthe far field, based on interpretation of seismic wav es.
→→ Difficult to measure true size of very large earthquakes.
→→ Uncertainties remain as to factors contributing to generationof tsunami.
Other Aspects
* I n the near field, essentially no time to issue warning.
Must rely on EDUCATION.
* Research must progress based on extremely small samples.
* I mplementation of warning procedures transcends Science.
* Other (newer) technologies may help, but are largelyunproven.
EARLY WARNING EFFORTS T.A. Jaggar, HVO
The first U.S. tsunami warning was issued by T.Jaggar upon recording of the Kamchatka earth-quake of 03 February 1923.
It was strictly on the basis of"very big and veryfar", but Jaggar, having computed distance fromS − P, correctly predicted the time of arrival ofthe wav e.
The warning was ignored; the tsunami did a lotof damage in Hawaii; one mariner was killed inHilo.
A new warning was later given on 02 March1933 (Sanriku earthquake), and this time nohuman lives were lost, despite significant flood-ing and damage.
In both instances, the earthquake had occurredin the morning, Hawaiian time.
Contrast with 1946, when at any rate, Jaggar had retired !
T. Jaggar at HVO,1920s
Damage on the Island of Hawai’i
03 February 1923
[Volcano Letter,1923]
DESTRUCTION OF THE LIGHTHOUSEAT SCOTCH CAP, UNIMAK Is.
[Photog.H. Hartman; Courtesy G. Fryer]
Before(1945)
After (est. 03-04 (?) Apr. 1946)
01 APRIL 1946 TSUNAMI in HILO, Hawaii
•
01 APR 1946: 1st CATASTROPHIC TSUNAMI in the U.S.
164 killed
MOTIV ATED CREATION ofPA CIFIC TSUNAMI WARNING CENTER
PTWC: The Pacific Tsunami Warning Center
Ewa Beach, Hawaii
Established 1949
Dr. Charles "Chip" McCreeryDirector
•PTWC
28 MAR 1964: ALASKA TSUNAMI, U.S. MAINLAND ATTA CKED
122 killedincluding 12
in California
MOTIV ATED CREATION of
ALASKA TSUNAMI WARNING CENTER
Seward, Alaska
Valdez, Alaska
Crescent City, California
[Ben-Menahem andRosenman,1972]
ATWC: The Alaska [West-Coast] Tsunami Warning Center
Palmer, Alaska
Established 1967
Dr. Paul WhitmoreDirector
•ATWC
TSUNAMI WARNING PROCEDURES at
Pacific Tsunami Warning Center, Ewa Beach, Hawaii
P T W C
2 scientists
ON SHIFT 24/7
∼ 100Seismic Stations(worldwide)
60MaregraphStations
Warning issued toPacific-rim countries
1. Detectearthquake3 minutes
2. Locateearthquake4 minutes} Automatic
3. Quantify earthquake / Assess risk:5 to 25 minutesScientist Input Critical
NOTE: Evacuationof Waikiki before tsunami would require 2.5 hours (HPD).
EARTHQ UAKES TAKE TIME T O OCCUR
• The larger the earthquake, the longer the source("Scaling Law").
• Measuring large earthquakes at small periods simply misses their true size.
• In the case of Sumatra, full size available only from normal modes.
mb Ms
TSUNAMITSUNAMI
GPS
TSUNAMI WARNING: THE CHALLENGE
EXTREME
PROBABLE
LOW
NIL
FAR-FIELDTSUNAMI DANGER
[Geller,1976]
Ms SATURATES AROUND 8.2
RECALL: ALL CONVENTIONAL MA GNITUDES
INAPPROPRIATE for TSUNAMI W ARNING
MmMm and TREMORS
[Okal and Talandier,1989]
NEWMagnitude Scale,Mm ,
using mantle Rayleigh wav es,with variablevariableperiod
• Directly related to seismic momentM0
• All constants justified theoretically
AUTOMATE PROCESS
* I mplemented,Papeete, Tahiti (1991),PTWC (1999)
MwP [Tsuboi,1996]
Idea: Try to recover the full moment information from thePwaves which arrive faster than the Rayleigh wav es.
• Note that formula forP waves inv olves
TIME DERIV ATIVE of MOMENT FUNCTION , XX
Idea is to computeTIME INTEGRALof P wave deformation torecover X, and hence static momentM0.
ALGORITHMS for MOMENT ESTIMATION IN USE A T PTWC
From [email protected] Mon Jun 28 05:06:18 2004Return-path: <[email protected]>Envelope-to: [email protected]: Mon, 28 Jun 2004 05:06:18 -0500Status: RO
TSUNAMI BULLETIN NUMBER 001PACIFIC TSUNAMI WARNING CENTER/NOAA/NWSISSUED AT 1000Z 28 JUN 2004
THIS BULLETIN IS FOR ALL AREAS OF THE PACIFIC BASIN EXCEPTALASKA - BRITISH COLUMBIA - WASHINGTON - OREGON - CALIFORNIA.
.................. TSUNAMI INFORMATION BULLETIN ..................
THIS MESSAGE IS FOR INFORMATION ONLY. THERE IS NO TSUNAMI WARNINGOR WATCH IN EFFECT.
AN EARTHQUAKE HAS OCCURRED WITH THESE PRELIMINARY PARAMETERS
ORIGIN TIME - 0950Z 28 JUN 2004 COORDINATES - 55.0 NORTH 133.5 WEST LOCATION - SOUTHEASTERN ALASKA MAGNITUDE - 7.0
EVALUATION
NO DESTRUCTIVE PACIFIC-WIDE TSUNAMI THREAT EXISTS BASED ON HISTORICAL EARTHQUAKE AND TSUNAMI DATA. HOWEVER - EARTHQUAKES OF THIS SIZE SOMETIMES GENERATE LOCAL TSUNAMIS THAT CAN BE DESTRUCTIVE ALONG COASTS LOCATED NEAR THE EARTHQUAKE EPICENTER. AUTHORITIES IN THE REGION OF THE EPICENTER SHOULD BE AWARE OF THIS POSSIBILITY.
THIS WILL BE THE ONLY BULLETIN ISSUED FOR THIS EVENT UNLESSADDITIONAL INFORMATION BECOMES AVAILABLE.
THE WEST COAST/ALASKA TSUNAMI WARNING CENTER WILL ISSUE BULLETINSFOR ALASKA - BRITISH COLUMBIA - WASHINGTON - OREGON - CALIFORNIA.
EXAMPLE of PTWC B ULLETIN
Benign -- No tsunami
+10 minutes
From [email protected] Wed Sep 12 06:31:32 2007Return-Path: <[email protected]>X-Spam-Checker-Version: SpamAssassin 3.1.7-deb (2006-10-05) on pluto.earth.northwestern.eduX-Spam-Level:X-Spam-Status: No, score=-1.3 required=4.5 tests=AWL,BAYES_00,UPPERCASE_50_75 autolearn=no version=3.1.7-debReceived: from infolist.nws.noaa.gov (infolist.nws.noaa.gov [140.90.19.83]) by pluto.earth.northwestern.edu (8.13.8/8.13.8/Debian-2) with SMTP idl8CBVUUg024717 for <[email protected]>; Wed, 12 Sep 2007 06:31:32 -0500Date: Wed, 12 Sep 2007 11:24:58 GMTFrom: PTWC <[email protected]>Message-Id:-220930-2007.09.12-07.31.02--emile#[email protected]> To: "ITIC Tsunami Bulletin Board" <[email protected]>Subject: Indian-Ocean-Wide Tsunami Watch BulletinList-Unsubscribe: <mailto:[email protected]>Reply-To: "ITIC Tsunami Bulletin Board" <[email protected]>Status: RO ITIC Tsunami Bulletin BoardTSUNAMI BULLETIN NUMBER 001PACIFIC TSUNAMI WARNING CENTER/NOAA/NWSISSUED AT 1124Z 12 SEP 2007 THIS BULLETIN IS FOR ALL AREAS OF THE INDIAN OCEAN. ... AN INDIAN-OCEAN-WIDE TSUNAMI WATCH IS IN EFFECT ... A TSUNAMI WATCH IS IN EFFECT FOR INDONESIA / AUSTRALIA / INDIA / SRI LANKA / THAILAND / UNITED KINGDOM / MALDIVES / MYANMAR / MALAYSIA / BANGLADESH / MAURITIUS / REUNION / SEYCHELLES / MADAGASCAR / SOMALIA / OMAN / PAKISTAN / IRAN / YEMEN / COMORES / CROZET ISLANDS / MOZAMBIQUE / KENYA / TANZANIA / KERGUELEN ISLANDS / SOUTH AFRICA / SINGAPORE THIS BULLETIN IS ISSUED AS ADVICE TO GOVERNMENT AGENCIES. ONLYNATIONAL AND LOCAL GOVERNMENT AGENCIES HAVE THE AUTHORITY TO MAKEDECISIONS REGARDING THE OFFICIAL STATE OF ALERT IN THEIR AREA ANDANY ACTIONS TO BE TAKEN IN RESPONSE. AN EARTHQUAKE HAS OCCURRED WITH THESE PRELIMINARY PARAMETERS ORIGIN TIME - 1110Z 12 SEP 2007 COORDINATES - 4.5 SOUTH 101.3 EAST LOCATION - SOUTHERN SUMATERA INDONESIA MAGNITUDE - 7.9 EVALUATION
EARTHQUAKES OF THIS SIZE HAVE THE POTENTIAL TO GENERATE A WIDESPREAD DESTRUCTIVE TSUNAMI THAT CAN AFFECT COASTLINES ACROSS THE ENTIRE INDIAN OCEAN BASIN. HOWEVER - IT IS NOT KNOWN THAT A TSUNAMI WAS GENERATED. THIS WATCH IS BASED ONLY ON THE EARTHQUAKE EVALUATION. AUTHORITIES IN THE REGION SHOULD TAKE APPROPRIATE ACTION IN RESPONSE TO THE POSSIBILITY OF A WIDESPREAD DESTRUCTIVE TSUNAMI. DUE TO ONLY LIMITED SEA LEVEL DATA FROM THE REGION IT MAY NOT BE POSSIBLE FOR THIS CENTER TO RAPIDLY CONFIRM NOR EVALUATE THE STRENGTH OF A TSUNAMI IF ONE HAS BEEN GENERATED. ESTIMATED INITIAL TSUNAMI WAVE ARRIVAL TIMES AT FORECAST POINTSWITHIN THE WARNING AND WATCH AREAS ARE GIVEN BELOW. ACTUALARRIVAL TIMES MAY DIFFER AND THE INITIAL WAVE MAY NOT BE THELARGEST. A TSUNAMI IS A SERIES OF WAVES AND THE TIME BETWEENSUCCESSIVE WAVES CAN BE FIVE MINUTES TO ONE HOUR. LOCATION FORECAST POINT COORDINATES ARRIVAL TIME -------------------------------- ------------ ------------ INDONESIA BENGKULU 3.9S 102.0E 1123Z 12 SEP SIBERUT 1.5S 98.7E 1203Z 12 SEP PADANG 0.9S 100.1E 1214Z 12 SEP BANDAR LAMPUNG 5.7S 105.3E 1242Z 12 SEP SIMEULUE 2.5N 96.0E 1243Z 12 SEP CILACAP 7.8S 108.9E 1307Z 12 SEP BANDA ACEH 5.5N 95.1E 1329Z 12 SEP BALI 8.7S 115.3E 1345Z 12 SEP KUPANG 10.0S 123.4E 1453Z 12 SEP BELAWAN 3.8N 99.0E 1703Z 12 SEP AUSTRALIA CHRISTMAS IS 10.4S 105.4E 1220Z 12 SEP COCOS ISLAND 12.1S 96.7E 1234Z 12 SEP
.........................................
IRAN GAVATER 25.0N 61.3E 1943Z 12 SEP YEMEN AL MUKALLA 14.5N 49.2E 2003Z 12 SEP ADEN 13.0N 45.2E 2100Z 12 SEP COMORES MORONI 11.6S 43.3E 2006Z 12 SEP CROZET ISLANDS CROZET ISLANDS 46.4S 51.8E 2009Z 12 SEP MOZAMBIQUE CABO DELGADO 10.7S 40.7E 2034Z 12 SEP ANGOCHE 15.5S 40.8E 2044Z 12 SEP
......................................... SOUTH AFRICA PRINCE EDWARD I 46.6S 37.6E 2146Z 12 SEP
DURBAN 29.8S 31.2E 2205Z 12 SEP PORT ELIZABETH 33.9S 25.8E 2256Z 12 SEP CAPE TOWN 34.1S 18.0E 2359Z 12 SEP THIS WILL BE THE ONLY BULLETIN ISSUED BY THE PACIFIC TSUNAMIWARNING CENTER FOR THIS EVENT UNLESS ADDITIONAL INFORMATIONBECOMES AVAILABLE. THE JAPAN METEOROLOGICAL AGENCY MAY ISSUE ADDITIONAL INFORMATIONFOR THIS EVENT. IN THE CASE OF CONFLICTING INFORMATION...THEMORE CONSERVATIVE INFORMATION SHOULD BE USED FOR SAFETY. ---
BENGKULU 12 SEP 2007 Message 1
+14 minutes
From [email protected] Wed Sep 12 09:52:29 2007Return-Path: <[email protected]>X-Spam-Checker-Version: SpamAssassin 3.1.7-deb (2006-10-05) on pluto.earth.northwestern.eduX-Spam-Level:X-Spam-Status: No, score=-1.3 required=4.5 tests=AWL,BAYES_00,UPPERCASE_50_75 autolearn=no version=3.1.7-deb
PACIFIC TSUNAMI WARNING CENTER/NOAA/NWSISSUED AT 1440Z 12 SEP 2007 THIS BULLETIN IS FOR ALL AREAS OF THE INDIAN OCEAN. ... AN INDIAN-OCEAN-WIDE TSUNAMI WATCH IS IN EFFECT ... A TSUNAMI WATCH IS IN EFFECT FOR INDONESIA / AUSTRALIA / INDIA / SRI LANKA / THAILAND / UNITED KINGDOM / MALDIVES / MYANMAR / MALAYSIA / BANGLADESH / MAURITIUS / REUNION / SEYCHELLES / MADAGASCAR / SOMALIA / OMAN / PAKISTAN / IRAN / YEMEN / COMORES / CROZET ISLANDS / MOZAMBIQUE / KENYA / TANZANIA / KERGUELEN ISLANDS / SOUTH AFRICA / SINGAPORE THIS BULLETIN IS ISSUED AS ADVICE TO GOVERNMENT AGENCIES. ONLYNATIONAL AND LOCAL GOVERNMENT AGENCIES HAVE THE AUTHORITY TO MAKEDECISIONS REGARDING THE OFFICIAL STATE OF ALERT IN THEIR AREA ANDANY ACTIONS TO BE TAKEN IN RESPONSE. AN EARTHQUAKE HAS OCCURRED WITH THESE PRELIMINARY PARAMETERS ORIGIN TIME - 1110Z 12 SEP 2007 COORDINATES - 4.5 SOUTH 101.3 EAST LOCATION - SOUTHERN SUMATERA INDONESIA MAGNITUDE - 8.2 MEASUREMENTS OR REPORTS OF TSUNAMI WAVE ACTIVITY GAUGE LOCATION LAT LON TIME AMPL PER ------------------- ----- ------ ----- --------------- ----- PADANG ID 0.9S 100.4E 1348Z 0.98M / 3.2FT 34MIN COCOS CC 12.1S 96.9E 1236Z 0.11M / 0.4FT 22MIN DART 23401 8.9N 88.5E 1421Z 0.02M / 0.1FT 30MIN LAT - LATITUDE (N-NORTH, S-SOUTH) LON - LONGITUDE (E-EAST, W-WEST) TIME - TIME OF THE MEASUREMENT (Z IS UTC IS GREENWICH TIME) AMPL - TSUNAMI AMPLITUDE MEASURED RELATIVE TO NORMAL SEA LEVEL. IT IS ...NOT... CREST-TO-TROUGH WAVE HEIGHT. VALUES ARE GIVEN IN BOTH METERS(M) AND FEET(FT). PER - PERIOD OF TIME IN MINUTES(MIN) FROM ONE WAVE TO THE NEXT. EVALUATION SEA LEVEL READINGS INDICATE A TSUNAMI WAS GENERATED. IT MAY
ALREADY HAVE BEEN DESTRUCTIVE ALONG SOME COASTS. BASED ON THESE DATA THE THREAT CONTINUES FOR ALL COASTAL AREAS OF THE INDIAN OCEAN. FOR THOSE AREAS - WHEN NO MAJOR WAVES HAVE OCCURRED FOR AT LEAST TWO HOURS AFTER THE ESTIMATED ARRIVAL TIME
OR DAMAGING WAVES HAVE NOT OCCURRED FOR AT LEAST TWO HOURS THEN LOCAL AUTHORITIES CAN ASSUME THE THREAT IS PASSED. DANGER TO BOATS AND COASTAL STRUCTURES CAN CONTINUE FOR SEVERAL HOURS DUE TO RAPID CURRENTS. AS LOCAL CONDITIONS CAN CAUSE A WIDE VARIATION IN TSUNAMI WAVE ACTION THE ALL CLEAR DETERMINATION MUST BE MADE BY LOCAL AUTHORITIES. DUE TO ONLY LIMITED SEA LEVEL DATA FROM THE REGION IT MAY NOT BE POSSIBLE FOR THIS CENTER TO RAPIDLY NOR ACCURATELY EVALUATE THE STRENGTH OF A TSUNAMI IF ONE HAS BEEN GENERATED.
BENGKULU 12 SEP 2007 Message 5
+ 03:30 hours
PROBLEMS WITH SURFACE−WAVE TECHNIQUES
( Mm )( Mm )
SURFACE WAVES ARESLOW ... THEY COME LATE
Computation need to be pushed to very long periods for Mega Events
(T ≥ 550 seconds).
The duration of the source may be so long that thePwave interferes with subsequent phases (PP, even S)
Example: Sumatra-AndamanEvent, 26 DEC 2004
Duration of Source: 500 to 600 seconds
Station MSEY (Mahe, Seychelles;∆ = 41°).
P S
M wp
Other Problems:
• Theory valid only infar-field
Yet, applied undiscriminately in both near- and far-fields
• Length of window / Frequency band never satisfacto-rily resolved
• Influence of depth phases / triplications not sorted out
• Operational details of algorithm unresolved
• Performance on large dataset, including tsunami earth-quakes, not assessed
• Empirical patches for big events (changeα h ??)unsatisfactory
• In time domain algorithm, instrument responsenot flat at long periods
PROBLEMS with BODY-WAVE TECHNIQ UES
For Mega - Events,
• The Cause:Earthquake has exceedingly slowrupture process releasing very little energy intohigh frequencies felt by humans and contributingto damage[Tanioka, 1997; Polet and Kanamori,2000].
• The Challenge: Can we recognize them fromtheir seismic wav es in [quasi-]real time?
• The Solution:The Θ parameter [Newman andOkal, 1998] compares the "size" of the earth-quake in two different frequency bands.
→ Use generalized−P wavetrain (P, pP, sP).
1994 Jav a"Tsunami Earthquake"
Station: TAU(Hobart, Tasmania)
→ DefineEstimated Energy, EE
EE = (1 + q)16
5
[a/g(15;∆)]2
(Fest)2ρ α
ωmax
ωmin
∫ ω2 u(ω) 2 eω t* (ω) ⋅ dω
→ Scale to Moment throughΘ = log10EE
M0
→ Scaling laws predictΘ = −4. 92.
• Tsunami earthquakes characterized byDeficientΘ (as much as 1.5 units).
EXAMPLE of COMPUT ATION of PARAMETER Θ
JAPAN
RED Events are SLOW (Θ ≤ − 5. 8)
Note that this event had a trend towards being FAST
(Outer ridge event NOT at SUBDUCTION INTERFACE)
DETECT "TSUNAMI EAR THQUAKES" in REAL TIME
NOW OPERATION AL at PTWC
MAN AGING a HUMONGOUS DAT ABASE
Weinstein and Okal[2005] studied 119,643 measurements ofMm at PTWC taken at 14 periods on 8546 seismograms from192 earthquakes recorded at 99 stations.
They examined various strategies to best determine the earthquake’s moment.
• USE ONLY STATIONS WITH
VERY-BROAD-BAND INSTRUMENTS
• DEVISE COMPLEX AVERAGING STRATEGIES
• DO NOT CLUSTER TOO MANY STATIONS
IN SAME AZIMUTH
PERU, 23 June 2001
WE FAILED
THAILAND8000 deaths
INDIA11000 deaths
SRI LANKA31000 deaths
SOMALIA300 deaths
SUMATRA 2004
SCIENCE did not FAIL; COMMUNICA TIONS DID.
To a large extent, the scientific processing of the 2004earthquake did not fail
Even though the final moment took one month to assess, avalue (8 to 9 times 1028 dyn-cm; Mw = 8. 5), sufficient totrigger a tsunami alert if the earthquake had been in thePacific Basin, was recognized in due time.
• COMMUNICATIONS INFRASTRUCTURE CANNOTBE IMPROVISED AND MUST BE DESIGNED, BUILTAND TESTED AHEAD OF TIME.
• The development of reliable tsunami systems in theAtlantic and Indian Oceans must focus on
COMMUNICATIONS,
to a greater extent than on additional seismic sensors.
• New observations (or the lack of data) point out to thepotential value of a synergy between various technologies.
SATURDAY
2 5DECEMBER 2004
CHRISTMAS
POST−MORTEM of a Failure
1. DEEP-OCEAN ASSESSMENT &REPORTING of TSUNAMIS
PMEL - NOAA, Seattle (E.N. Bernard; F.I. Gonzalez; H.B. Milburn)
• Use pressure sensor at the bottom of theocean to detect the overpressure causedby the passage of the tsunami.
• Relay information by satellitethrough buoy.
NEW TECHNOLOGIES: Directly detect TSUNAMI rather than Earthquake !
160˚
160˚
170˚
170˚
180˚
180˚
190˚
190˚
200˚
200˚
210˚
210˚
45˚ 45˚
50˚ 50˚
55˚ 55˚
60˚ 60˚
D-171
ALEUTIAN — 17 NOV 2003
First successful operational use of DART sensors to call off an alert.
This signal was interpreted by V.V. Titovin terms of source size and a real-timesimulation performed to predict the run-up in Hawaii. As a result, the pendingalert was called off by PTWC.
DART
A Growing, High-Performing Network
SUCCESSFUL OPERATIONAL USE
Rayleigh Tsunami
Problems:
• COST
• Vandalism
TOWARDS DIRECT DETECTION of a TSUNAMI on the HIGH SEAS2. TSUNAMI DETECTION by SATELLITE AL TIMETRY
DETECTION by SATELLITE AL TIMETRY gives firstdefinitive measurement ofMAJORtsunami onHIGH SEAS
Measures 70 cm across Bay of Bengal
cm
75˚ 90˚ 105˚
-15˚
0˚
15˚
02:52
03:02
00:59
SUMATRA 2004 →
Promise: WEAK
THE SATELLITE NEEDS TO BE AT THE RIGHT PLACE AT THE RIGHT TIME!!
TOWARDS DIRECT DETECTION of a TSUNAMI on the HIGH SEAS3. TSUNAMI DETECTION by GPS IONOSPHERIC MONIT ORINGJ. Artru, H. Kanamori (Caltech);M. Murakami (Tsukuba); P. Lognonne´, V. Du`cic (IPG Paris) -- (2002)
• Ocean surface is not free boundary — Atmosphere has finite density
• Tsunami wav eprolongedinto atmosphere;amplitude increaseswith height.
• Perturbation in ionosphere (h = 150−350 km) detectable by GPS.
Gravity Wav eProlongingTsunamiUpwards
Amplitude: 0.1 − 1 km
Amplitude: 10 cm
28 MAR 2000 -- 90 mn after earthquake
SUMATRA 2004Perturbations detected in ionosphericTotalElectronContent [Liu et al.,2006]
Successfully modeled byOcchipinti et al.[2006].PreviousDay
TSUNAMI
ELLIPTICITY of TSUNAMI SURF ACE MOTION
(Shallow Water Approximation)
ux
uz=
1
ω √ g
h= typically = 10 to 30
Sumatra 2004:uz ≈ 1 m (JASON; seismic stations)
ux ≈ 15 meters ?
Conceivable to use GPS-equipped ships to detect tsunami.
TsunamiTsunami
Ship A should see a perturbation in speed
Ship B would show a zig-zag in trajectory
